Indonesian Journal of Electrical Engineering and Computer Science
Not a member yet
9109 research outputs found
Sort by
Transforming E-governance: the potential of blockchain in the public sector
Get PDFBlockchain technology has become a transformative innovation in the digital governance landscape, offering new opportunities to enhance transparency, accountability, and citizen trust. This study offers an extensive bibliometric and thematic examination of international research on blockchain in E-governance from 2019 to 2024. Using data from the Scopus database, the analysis examines publication trends, leading countries, collaboration networks, and the intellectual structure of the field. The findings reveal that research output has grown steadily, dominated by technologically advanced nations such as China, India, and the United Kingdom. The thematic mapping identifies core clusters, including transparency, E-government, and public sector innovation, alongside emerging themes such as artificial intelligence (AI) integration, smart cities, and digital transformation. By integrating bibliometric and thematic analyses, this study offers a comprehensive understanding of how blockchain research evolves within public governance. Despite significant progress, challenges remain, particularly regarding empirical validation, governance frameworks, and regional disparities in adoption. Future research should explore a more complex roadmap for blockchain implementation in government through three interrelated dimensions: technical development, policy and regulatory frameworks, and socio-institutional adaptation. This multidimensional perspective underscores blockchain’s capacity to support secure, inclusive, and data-driven forms of digital governance
Aspect based multimodal sentiment analysis of product reviews using deep learning techniques
Get PDFSentiment analysis plays a crucial role in understanding customer opinions, particularly in product reviews. Traditional approaches primarily focus on textual data; however, with the rise of social media, incorporating multimodal data, including text and emojis, enhances sentiment analysis accuracy. This research introduces a multimodal aspect-based sentiment analysis (MABSA) framework, integrating textual and emoji representations for Samsung M21 product reviews from Flipkart. The methodology involves data preprocessing, aspect extraction, sentiment grouping, and feature extraction using deep learning (DL) techniques. Bidirectional long shortterm memory (Bi-LSTM) networks are employed for classification, leveraging Word2Vec, Emoji2Vec, and bidirectional encoder representations from transformers (BERT) embeddings. Experimental results show that BERT with Bi-LSTM outperforms Word2Vec with Bi-LSTM, achieving 95.6% accuracy in aspect prediction and 96.28% accuracy in sentiment classification. Comparative analysis with existing models highlights the superiority of the MASAT model, effectively integrating implicit aspects, emoticons, and emojis. The study demonstrates the importance of multimodal sentiment analysis for a more comprehensive understanding of user opinions, offering valuable insights for businesses to enhance customer satisfaction
Evaluating multilingual encoder models for few-shot named entity recognition tasks
Get PDFThis work provides a thorough analysis of few-shot learning approaches in the realm of multilingual named entity recognition (NER). Our research is driven by the need to enhance linguistic inclusivity and performance efficiency across diverse languages. We focus on benchmarking a selection of prominent encoder models including XLM-RoBERTa (XLM-R), multilingual BERT (mBERT), DistilBERT, character architecture for eNcoders IN embeddings (CANINE), and multilingual text-to-text transfer transformer (mT5), to illuminate their capabilities and limitations within few-shot learning paradigms, particularly for underrepresented languages. Results indicate that models like XLM-R and mT5 demonstrate superior adaptability and accuracy, outperforming others in complex linguistic settings, which suggests their potential in supporting more inclusive artificial intelligence (AI) technologies. The impact of this study extends beyond academic interest, offering pivotal insights for the development of more inclusive, adaptable and efficient NER systems. By advancing our understanding of few-shot learning in multilingual contexts, this work contributes to the broader goal of creating AI applications that are linguistically diverse and more reflective of global communication patterns. These results provide crucial insights for advancing entity recognition capabilities across diverse artificial intelligence systems, facilitating development of more precise, equitable, and sophisticated linguistic processing frameworks
Dynamic resource allocation in cloud-radio access network using call detail record analysis
Get PDFWe propose a solution based on call detail record (CDR) data analysis for cloud-radio access network (C-RAN) network optimization. First, we propose a data traffic prediction model in 3G and 4G networks using artificial intelligence (AI) models (long short-term memory (LSTM) and Bidirectional LTSM (BiLSTM)). Second, we propose a dynamic baseband units (BBU) resource allocation algorithm based on the obtained traffic prediction results to evaluate the rate of BBUs used as well as the average utilization rate of active BBUs in a C-RAN network. We used mean absolute error, root mean square error and mean absolute percentage error to evaluate the prediction model. The results obtained show that the best performance for estimating data traffic in 3G and 4G networks was obtained with the BiLSTM model, and is as follows: 1.143; 1.521; 2.47 percent for 4G, and for 3G, we have 0.2553; 0.3608 and 27.70 percent. Finally, evaluations with the predicted traffic dataset show that our framework provides up to 81% reduction in the number of BBUs used by the normal RAN. Moreover, active BBUs are exploited on average up to 88.34% of their capacity in a C-RAN compared to an average rate of 10.8% in a traditional RAN
Deep-learning-based hand gestures recognition applications for game controls
Get PDFHand gesture recognition is among the emerging technologies of human computer interaction, and an intuitive and natural interface is more preferable for such applications than a total solution. It is also widely used in multimedia applications. In this paper, a deep learning-based hand gesture recognition sys tem for controlling games is presented, showcasing its significant contributions toward advancing the frontier of natural and intuitive human-computer interac tion. It utilizes MediaPipe to get real-time skeletal information of hand land marks and translates the gestures of the user into smooth control signals through an optimized artificial neural network (ANN) that is tailored for reduced com putational expenses and quicker inference. The proposed model, which was trained on a carefully selected dataset of four gesture classes under different lighting and viewing conditions, shows very good generalization performance and robustness. It gives a recognition rate of 99.92% with much fewer param eters than deeper models such as ResNet50 and VGG16. By achieving high accuracy, computational speed, and low latency, this work addresses some of the most important challenges in gesture recognition and opens the way for new applications in gaming, virtual reality, and other interactive fields
Experimental analysis and bug abstraction for distributed computation on ray framework
Get PDFThis research aims to address challenges in distributed computing, focusing on the ray framework, which has potential for efficient parallel and distributed task execution. While methods such as model-checkers and fuzzing have been applied to detect bugs, both have limitations in handling the complexity of distributed computing, particularly in dealing with issues like state-space explosion and identifying rare bugs. This study proposes an alternative approach through experimental analysis and bug abstraction methods to discover, identify, and classify bugs in the ray framework. Experimental analysis involves isolating and re-testing bugs in a controlled environment to understand their characteristics, while bug abstraction analyzes the factors causing bugs to identify common patterns and characteristics. The results of this research successfully identified three main categories of bugs: crash, performance, and inaccurate status, and revealed bug characteristics that do not depend on actor instance multiplicity, actor type, specific event sequences, or particular configurations. This research makes a significant contribution to the development of more effective and efficient bug detection methods in distributed computing, particularly in the ray framework, and paves the way for further research to enhance the reliability of distributed systems.
Detection of short circuit faults in two-level voltage source inverter using convolution neural network
Get PDFVoltage source inverters (VSIs) play a critical role in modern industrial systems, particularly in controlling the operation of equipment such as induction motors. Ensuring their reliable performance is crucial, as faults like short circuits can severely disrupt industrial processes. This paper introduces a new diagnostic approach for detecting and localizing short circuit faults in VSIs. The method uses Lissajous curves derived from the Clark transformation of the VSI’s 3-phase voltage components (Vα, Vβ). These curves serve as input data for a convolutional neural networks (CNNs) model, enabling the accurate classification of single and double short circuit faults. Simulation results using MATLAB/Simulink demonstrate that the proposed method achieves 100% classification accuracy within 100 ms, highlighting its suitability for real-time applications. The approach offers significant advantages in speed and accuracy over traditional techniques, with potential implications for enhancing the reliability and safety of inverter-driven systems in industrial environments
Meta-model integration with attention mechanisms for advanced decision-level fusion in machine learning
Get PDFThis work proposes an advanced meta-model approach that incorporates forecasts from multiple machine learning models to improve classification accuracy in complex tasks. The approach employs decision-level data fusion, where predictions from random forest (RF), XGBoost, neural networks (NN), and support vector machine (SVM) are combined within a meta-model framework. The meta-model incorporates an attention mechanism and a gated model selection process to dynamically emphasize the most relevant model outputs based on input features. The results demonstrate superior accuracy in predicting explicit content compared to traditional fusion methods. This research highlights the potential of attention-enhanced meta-models in improving interpretability and accuracy across various domains. The integration of meta-models with attention mechanisms has the potential to significantly enhance decision-level fusion in machine learning applications. This study investigates the development of an advanced fusion framework leveraging attention mechanisms to improve decision-making accuracy in multi-source data environments. The proposed method is evaluated across multiple datasets, demonstrating its efficacy in increasing predictive performance and robustness
Random forest algorithm with hill climbing algorithm to improve intrusion detection at endpoint and network
Get PDFCloud computing is a framework that enables end users to connect highly effective services and applications over the internet effortlessly. In the world of cloud computing, it is a critical problem to deliver services that are both safe and dependable. The best way to lessen the damage caused by entry into this environment is one of the primary security concerns. The fundamental advantage of a cooperative approach to intrusion detection system (IDS) is a superior vision of an action of network attack. This paper proposes a random forest (RF) algorithm with a hill-climbing algorithm (RFHC) to improve intrusion detection at the endpoint and network. Initially, it is used for feature selection, and the next process is to separate the intrusions detection. The feature selection is maintained by the hill climbing (HC) algorithm that chooses the best features. Then, we utilize the RF algorithm to separate the intrusion efficiently. The experimental results depict that the RFHC mechanism reached more acceptable results regarding recall, precision, and accuracy than a baseline mechanism. Moreover, it minimizes the miss detection ratio and enhances the intrusion detection ratio
Application of data mining for diagnosis of ENT diseases using the Naïve Bayes method with genetic algorithm feature selection
Get PDFEar, nose, and throat (ENT) disease is a disorder that occurs in the eustachian tube in one of the organs, be it the ear, nose, or throat. Early signs of ENT disease include sore throat, painful swallowing, swollen and red tonsils, runny nose, nosebleeds, blocked nose, discharge from the ears, and others. To determine the diagnosis, it is necessary to carry out a physical examination of the ears, nose, and throat as recommended by an expert, namely an ENT doctor. The research carried out was implementing data mining for the diagnosis of ENT diseases using the Naïve Bayes (NB) method. This method was chosen because it can increase the accuracy, efficiency, and accessibility of health services and is also easy to understand and apply to classify ENT disease symptom data. The NB method was used to build an ENT diagnosis classification model and the model performance was evaluated using accuracy, precision, and recall metrics. To increase the accuracy of the NB algorithm predictions, feature selection using a genetic algorithm can be used. Genetic algorithms can help select the most relevant and significant features, improving the accuracy of NB models by eliminating irrelevant or noisy features. By applying this method, predictions for ENT diseases can be produced with an accuracy of 95.67%